Medical Marijuana and Cannabidiol (CBD)

In the last few years there has been renewed interest in the therapeutic potential of marijuana (cannabis and its derivatives) for myriad conditions including epilepsy, and specifically infantile spasms. The overwhelming source of enthusiasm has been from online parent testimonials and press reports of success in individual patients. Perhaps the most balanced parent report was published in Wired magazine.  Like many "alternative" therapies which are outside mainstream medicine, there is limited—but growing—scientific support for the use of cannabis to treat epilepsy. Although the internet is replete with testimonials extolling the virtue of medical marijuana, there is insufficient evidence—and especially safety data—to recommend its use for treatment of seizures in most cases. Please see our discussion of "Treatment X". However, for patients who have not responded to first-line mainstream therapies, and for whom the potential benefit and risks of ongoing seizures/spasms outweigh the largely unknown risks of marijuana therapy, a marijuana preparation may be reasonable. This should take place only after a candid and thorough conversation with a neurologist who is familiar with these therapies and associated risk. 

It is dangerous to try these therapies without informing your child's physicians, especially those who prescribe other drugs. Most readily available formulations of medical marijuana contain numerous—more than 100—chemicals derived from the plant, many of which may interact with commonly prescribed medications. There is compelling evidence of a strong interaction between cannabidiol (CBD) and clobazam (Onfi®, Frizium®), such that blood levels of an active metabolite (N-desmethyclobazam, or norclobazam) may rise considerably. Conversely, clobazam exposure may increase blood levels of cannabidiol. This effect on clobazam metabolism may explain favorable response in some patients, as well as some side effects that have been attributed to cannabidiol. 

In most countries, there are many indirect and legal risks that accompany the use of marijuana preparations obtained in the community without oversight by regulatory agencies such as the Food and Drug Administration (FDA) and Drug Enforcement Agency (DEA). In the United States, although medicinal and recreational marijuana use is now legal in many states, marijuana continues to be classified as a "schedule I" controlled substance and is illegal according to federal law. Therefore, children, parents, "prescribing" physicians, and "manufacturers" (growers/distributors) of marijuana preparations in all states are potentially vulnerable to arrest and federal prosecution. Furthermore, the possession and administration of these products are officially forbidden in most hospitals because of DEA oversight and licensure. Having said this, we are unaware of any case in the United States in which a parent, health-care provider, or hospital has been prosecuted or faced other disciplinary action for using, administering, or recommending a marijuana-derived treatment in a good-faith effort to treat epilepsy.

There are efforts underway to rigorously evaluate the potential usefulness of medical marijuana for treatment of epilepsy in children. Specifically, several active drugs derived from marijuana, especially CBD, are being investigated as a potential treatments of severe epilepsy in children, including infantile spasms at UCLA [a study conducted by authors of this website]. A purified plant-derived form of CBD is being developed and tested by GW Pharmaceuticals, and a pure synthetic form of CBD is being developed and tested by Insys Therapeutics. GW Pharmaceuticals recently announced a successful randomized, double-blind, placebo-controlled studies of CBD for treatment of Dravet syndrome and Lennox Gastaut syndrome, though results have not yet appeared in a peer-reviewed scientific publication. CBD contrasts with the principal psychoactive component of marijuana called tetrahydrocannabinol (THC). THC is thought to be the constituent most responsible for the "high" (euphoria) of marijuana and associated effects on appetite and cognition. THC may be "proconvulsant" in some settings, meaning that consumption of THC may provoke seizures. In a parallel effort—not regulated by the FDA—several manufacturers (growers) have been breeding plants with high CBD content and low THC content. CBD, and other cannabinoids including THC, can be extracted from these plants using a variety of methods to produce a liquid/oil preparation which can be consumed. Despite rather compelling and emotional reports in the media, there continue to be major concerns regarding effectiveness, safety, and quality control (esp. CBD/THC content, variation in potency/concentration from batch to batch, shelf-life, pesticide content, and infectious—especially fungus—contamination). The FDA recently published a series of reports in 20152016, and 2017 in which they found that numerous community CBD products were inappropriately labeled with regard to CBD content—in some cases with egregious discrepancies, and that multiple manufacturers have made unfounded claims of effectiveness and safety.

The potential value of CBD-enriched cannabis extracts and pharmaceutical CBD preparations has been a focal point in recent meetings of the American Epilepsy Society. A few highlights are summarized here: 

Potential efficacy of cannabidiol for treatment of refractory infantile spasms and Lennox-Gastaut syndrome

Authors: Raymond Zhou, Catherine Jacobson, Julius Weng, Emily Cheng, Johnson Lay, Phoebe Hung, Jason Lerner, Raman Sankar and Shaun Hussain

This study—conducted by UCLA investigators who manage this website—is a presentation of an online survey of parents who have administered CBD-enriched cannabis preparations to their children for the treatment of severe epilepsy, including many patients with infantile spasms. Among 200 respondents, 117 had actually administered CBD products to their children. 53 children suffered from infantile spasms and/or Lennox Gastaut syndrome. Over 90% reported reductions in seizure frequency, and 13% reported seizure-freedom. Reported side effects were modest. Among the minority of parents who were able and willing to report exact CBD dosage, the typical dose was 4 mg/kg/day (1.8 mg/pound/day). Dosage was not different among children with and without improvement.

Cannabidiol treatment of refractory epileptic spasms: an open label study

Authors: Elena Abati, Evan Hess, Amy Morgan, Patricia L. Bruno, Elizabeth Thiele

In this report, Dr. Abati and colleagues report on the response of 9 patients with epileptic spasms (infantile spasms persisting well into childhood) to purified cannabidiol in a small open-label study. 30-50% of children had at least a 50% reduction in burden of seizures at 3 to 12 months, and 2 children were seizure-free at 2 months.

Parental reporting of response to oral cannabis extracts as adjunctive treatment for medically refractory epilepsy

Authors: Craig Press, Kelly Knupp and Kevin Chapman

In this report, Drs. Press, Knupp, and Chapman reviewed the medical records of 58 patients (mostly children) cared for at Children’s Hospital Colorado who received cannabis extracts for treatment of epilepsy. 48% reported improvement in seizures and 31% reported at least 50% reduction in seizure-frequency. Curiously, the reported response rate among children who had moved to Colorado to obtain cannabis extracts was far higher than reported response rates for children already living in Colorado (52% vs. 17%). Dosage data were not available.

A phase-2 study of pharmaceutical grade synthetic cannabidiol oral solution for treatment of refractory infantile spasms

Authors: Shaun Hussain, Dennis Dlugos, M. Roberta Cilio, Neha Parikh, Alexander Oh, Raman Sankar

In this study Dr. Hussain and colleagues treated 9 patients with infantile spasms who had previously failed treatment with ACTH and vigabatrin. Although 8 of 9 patients exhibited no response, one patient exhibited rapid resolution of both hypsarrhythmia and seizures (infantile spasms). This sole responder relapsed shortly thereafter. Given the severity and duration of infantile spasms in this cohort, this modest result suggests that cannabidiol may be effective, especially in less refractory populations. Further study is clearly warranted, and this study certainly does not prove that cannabidiol is (or is not) effective and safe in the treatment of infantile spasms.


These studies do not provide adequate evidence to conclude that pure CBD or CBD-enriched cannabis extracts are safe and effective. All investigators highlighted the need for well-designed clinical trials to properly evaluate the effectiveness and safety of these products. These studies demonstrated wide discrepancies in response rate, and none were able to control for the impact of bias (e.g. placebo-like effects). Despite the favorable response rate in the study and Zhou and colleagues, it is still quite possible that CBD is ineffective and unsafe. The fantastically high response rates observed by Zhou et al and the observation of Press et al that patients who moved to Colorado were substantially more likely to report benefit suggests that response rates are highly vulnerable to bias; at this point all response rates to date should be interpreted with extreme caution.


Although efforts are made to keep this website correct and up-to-date, we urge caution in interpreting the information you find here. This is in no way a substitute for the advice and care of a pediatric neurologist. Please view the terms of use.

English | Español